CN112434072A

CN112434072A - Searching method, searching device, electronic equipment and storage medium

Info

Publication number: CN112434072A
Application number: CN202110107813.0A
Authority: CN
Inventors: 桑梓森; 苑爱泉; 徐花; 芦亚飞; 何旺贵; 万家雪; 朱培源; 马骐; 许林隆; 王宇昊
Original assignee: Zhejiang Koubei Network Technology Co Ltd
Current assignee: Zhejiang Koubei Network Technology Co Ltd
Priority date: 2021-01-27
Filing date: 2021-01-27
Publication date: 2021-03-02
Anticipated expiration: 2041-01-27
Also published as: CN112434072B

Abstract

The application relates to the technical field of data search, and discloses a search method, a search device, electronic equipment and a storage medium, wherein the search method comprises the following steps: acquiring a search request of a user and candidate search results corresponding to the search request; determining a search intent category of the search request; determining the correlation degree of the search request and each candidate search result according to a correlation degree determination mode corresponding to the search intention category; and determining a target search result of the user from each candidate search result and providing the target search result to the user based on the corresponding correlation of each candidate search result. By the method, the accuracy of the search result can be improved.

Description

Searching method, searching device, electronic equipment and storage medium

Technical Field

The present application relates to the field of data search technologies, and in particular, to a search method, an apparatus, an electronic device, and a storage medium.

Background

In the search field, a user can inquire and search information by inputting search information, and in an application program, the user can search information by inputting the search information in a search box in the application program, and the search is an important way for the application program to acquire the user requirement and perform data interaction. The application degree background usually performs corresponding search according to the search condition input by the user to obtain a search result and returns the search result to the user. However, in practical situations, the meanings of the search conditions input by the user to be expressed are complex, and the same search conditions may express different meanings, so that the accuracy of the search results obtained based on the search information is not high, and the user expectations cannot be met.

Disclosure of Invention

The present application aims to solve at least one of the above technical drawbacks, and particularly proposes the following technical solutions to solve the problem of low accuracy of the search result.

In one aspect of the present application, a search method is provided, including:

acquiring a search request of a user and candidate search results corresponding to the search request;

determining a search intention category of the search request;

determining the correlation degree of the search request and each candidate search result according to the correlation degree determination mode corresponding to the search intention category;

and determining a target search result of the user from the candidate search results and providing the target search result to the user based on the corresponding correlation degree of the candidate search results.

In another aspect of the present application, there is provided a search apparatus, including:

the candidate search result acquisition module is used for acquiring a search request of a user and each candidate search result corresponding to the search request;

a search intention category determination module for determining a search intention category of the search request;

the relevancy determining module is used for determining the relevancy between the search request and each candidate search result according to a relevancy determining mode corresponding to the search intention category;

and the target search result determining module is used for determining the target search result of the user from the candidate search results and providing the target search result to the user based on the corresponding correlation of the candidate search results.

In yet another aspect of the present application, an electronic device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the searching method shown in the first aspect of the present application is implemented.

In yet another aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored, which when executed by a processor implements the search method shown in the first aspect of the present application.

The beneficial effect that technical scheme that this application provided brought is:

according to the searching method, the searching request is divided into the searching intention categories, the relevance determining mode corresponding to the searching intention categories of the searching request is determined, the relevance calculation is performed according to different searching intention categories, the relevance between the searching request and each candidate searching result is favorably and accurately obtained according to the searching intention of the user, the target searching result provided for the user is enabled to be more accordant with the expectation of the user based on the accurate relevance corresponding to each candidate searching result, the matching degree of the target searching result and the expectation of the user is improved, and the use perception of the user is improved.

Additional aspects and advantages of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the present application.

Drawings

The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

FIG. 1 is a flow chart of a search method provided by an embodiment of the present application;

FIG. 2 is a flowchart of a search method according to another embodiment of the present application, which shows a flowchart of a search for proper nouns (right part) and a flowchart of a search for non-proper nouns (left part);

FIG. 3 is a flowchart of determining a text match between each candidate search result and a search request according to an embodiment of the present application;

fig. 4 is a flowchart of determining a matching degree between related information corresponding to a candidate search result and related information of a search request according to an embodiment of the present application;

FIG. 5 is a diagram illustrating ranking of candidate search results based on relevance according to an embodiment of the present application;

fig. 6 is a schematic structural diagram of a search apparatus according to an embodiment of the present application;

fig. 7 is a schematic structural diagram of an electronic device according to an embodiment of the present application.

Detailed Description

Reference will now be made in detail to the embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are exemplary only for the purpose of explaining the present application and are not to be construed as limiting the present application.

As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may also be present. Further, "connected" or "coupled" as used herein may include wirelessly connected or wirelessly coupled. As used herein, the term "and/or" includes all or any element and all combinations of one or more of the associated listed items.

It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the prior art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

The term "person" refers to a specific or unique person or thing, such as: name of person, place, country, landscape, brand, address, shop, etc.

The relevance is the matching degree between the search request input by the user and the returned result, and the higher the relevance is, the higher the matching degree is, and the more the returned result is in line with the expectation of the search request of the user.

The scheme provided by the embodiment of the application can be executed by any electronic device, such as a terminal device, or a server, wherein the server can be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing cloud computing service. The terminal may be, but is not limited to, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and the like. The terminal and the server may be directly or indirectly connected through wired or wireless communication, and the application is not limited herein. For technical problems existing in the prior art, the search method, the search device, the electronic device and the storage medium provided by the application aim to solve at least one of the technical problems in the prior art.

The following describes the technical solutions of the present application and how to solve the above technical problems in detail with specific embodiments. The following several specific embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments. Embodiments of the present application will be described below with reference to the accompanying drawings.

The embodiment of the present application provides a possible implementation manner, and as shown in fig. 1, a flowchart of a search method is provided, where the scheme may be executed by any electronic device, and optionally may be executed at a server side or a terminal device, and for convenience of description, the method provided in the embodiment of the present application is described below with a server as an execution subject. As shown in fig. 1, the method may include the steps of:

step S110, obtaining a search request of a user and candidate search results corresponding to the search request;

step S120, determining the search intention category of the search request;

step S130, determining the correlation degree of the search request and each candidate search result according to the correlation degree determining mode corresponding to the search intention category;

and step S140, determining a target search result of the user from the candidate search results and providing the target search result to the user based on the corresponding correlation of the candidate search results.

The scheme provided by the application can be applied to, but is not limited to, the following scenes: the electronic device (e.g., a server) obtains a search request of a user, where the search request may be in the form of text, voice, or the like, and performs a preliminary search using the search request to obtain a plurality of candidate search results, where the search request is a product, and the candidate search results may be related products, stores containing the product, or the like. Then, the search intention category of the search request is identified and divided according to whether the search request contains proper nouns or not, and the classification can also be realized by an intention identification model, the intention recognition model can be trained based on a large amount of training data labeled with search intention, the intention recognition model can implement binary classification of search requests, that is, the search intention categories may be divided into proper noun searches and non-proper noun searches, since it is highly likely that the search results that each search intention category wishes to feed back are also different, therefore, the method provided by the embodiment of the application is provided with a corresponding relevance determining mode correspondingly for each type of search intention category, and determines the relevance between the search request and each candidate search result by using the relevance determining mode corresponding to the search intention category, wherein the relevance can be measured in a score, a level and the like. And determining a target search result which is in line with the expectation of the user according to the corresponding correlation degree of each candidate search result, and sending the target search result to the user.

Taking the search request as the commodity search request as an example, the candidate search results corresponding to the search request may include not only similar commodities but also shops hosting the commodities, that is, the candidate search results may include not only the same or similar commodities but also shops having a matching relationship with the commodities (for example, the shops sell the commodities), so that the range of the candidate search results is expanded, and the probability that the candidate search results meet the expectation of the user is improved.

And the degree of correlation between the search request and the candidate search result is the matching degree between the search request and the candidate search result. The higher the degree of correlation is, the more the candidate search result conforms to the expectation of sending a search request by a user; the lower the relevance, the lower the degree of match of the search request with the candidate search results, and the less desirable it is for the user.

In alternative embodiments of the present application, the search intention category includes proper noun search or non-proper noun search.

Proper noun search means that a user includes a proper noun, or a variant or alternative name of the proper noun (i.e. a possible proper noun, which may also be called a candidate proper noun) in an input search request, for example, the search request includes: the computer, the noun has specific meaning, for this kind of word that has specific meaning, utilize its corresponding relevance confirming mode to confirm the degree of correlation of search request and candidate search result, through carrying out classification to the search request, and utilize the degree of correlation confirming mode of pertinence to confirm the degree of correlation, be favorable to promoting the precision of the degree of correlation of search request and candidate search result.

The non-specific noun search is a search for words without specific meaning, such as category words, content words, and the like, for example, the search request includes a search for coffee, and the corresponding candidate search results may include vanilla coffee, ice cream coffee, latte, coffee shop a, western fast food shop B, sports gym D, and the like. The search types corresponding to the non-specific nouns are complex, and for the candidate search results of the relatively complex search requests, the correlation between the search requests and the candidate search results is determined by adopting a correlation determination mode aiming at the search intention category, so that the accuracy of the correlation is favorably improved.

According to the searching method, the searching request is divided into the searching intention categories, the relevance determining mode corresponding to the searching intention categories of the searching request is determined, the relevance calculation is performed according to different searching intention categories, the relevance between the searching request and each candidate searching result can be obtained accurately, the target searching result provided for the user can better accord with the expectation of the user based on the accurate relevance corresponding to each candidate searching result, and the matching degree of the target searching result and the expectation of the user is improved.

In order to make the searching scheme and its technical effect provided by the present application clearer, the following describes in detail a specific implementation of the searching scheme with a plurality of alternative embodiments.

In an alternative embodiment, the determining the relevance of the search request and each candidate search result according to the relevance determining manner corresponding to the search intention category provided in step S130 above may include:

a1, mining candidate proper nouns in the search request;

a2, determining the standard proper nouns corresponding to the candidate proper nouns based on the matching degree of the candidate proper nouns and the standard proper nouns in the proper noun database;

a3, based on the matching degree of the standard proper noun corresponding to the candidate proper noun and each candidate search result, determining the correlation degree between the search request and each candidate search result.

The proper noun search category may include other non-proper nouns besides proper nouns in the corresponding search request, or only include a part of proper nouns in the search request, by mining the search request, such as: the text in the search request is subjected to structured parsing, supplementation and the like to obtain candidate proper nouns, for example: the search request input by the user is a search request of 'ABCDEF', the brand name stored in the proper noun database is 'ABC', and the search request is structurally analyzed to obtain candidate proper nouns, such as 'AB', 'ABC', 'ABCDE', 'ABCDEF', and the like.

The matching degree of the candidate proper nouns and the standard proper nouns stored in the proper noun database is obtained, the matching degree can be represented by the text matching degree, the standard proper nouns with the highest matching degree can be determined as the standard proper nouns corresponding to the candidate proper nouns, and a large number of standard proper nouns and candidate inputs with the mapping relation with the standard proper nouns are stored in the proper noun database in advance. The candidate inputs can be obtained through dictionary prediction and the like, or can be obtained through historical data, and a plurality of candidate inputs are mapped with corresponding standard proper nouns in advance and stored in a proper noun database. By matching candidate proper nouns in the search request with candidate entries, standard proper nouns, stored in a proper noun database. The matching degree between each candidate proper noun and each standard proper noun is obtained, and the standard proper noun with the highest matching degree can be used as the standard proper noun corresponding to the candidate proper noun.

And when the matching degree of the candidate proper noun and the candidate input in the proper noun data is the highest, taking the standard proper noun which is associated with the candidate input in the proper noun data as the standard proper noun corresponding to the candidate proper noun.

And calculating the matching degree between the standard proper nouns corresponding to the candidate proper nouns and each candidate search result, and taking the matching degree as the correlation degree between the search request and each candidate search result.

An alternative embodiment provides a flow chart of the searching method as shown in fig. 2, which includes a flow chart of searching for proper nouns (right part in fig. 2) and a flow chart of searching for non-proper nouns (left part in fig. 2), and the searching process for proper nouns searching categories is described in conjunction with the right part in fig. 2: when a search request is received, and the search intention category of the search request is determined to be proper noun search, the proper noun may be a brand word or an address word (for example, administrative area, road, business district, market, etc.), for the search request including the part of proper nouns, data mining is performed on the search request to obtain candidate proper nouns included in the search request, then standard proper nouns stored in a proper noun database are obtained through dictionary prediction, then accurate matching is performed by using the standard proper nouns corresponding to the search request and each candidate search result, the accurate matching may be performed in a text matching manner or an entity information matching manner, for example, a user searches a "C-face museum", and after the dictionary prediction, the "C-face museum" and the "C-old face museum" are stored in the proper noun database, and the accurate matching may be performed in a text matching manner, and finally, taking the matching degree corresponding to the matching result of the accurate matching as the correlation degree of the search request and the corresponding candidate search result.

According to the embodiment of the application, the relevance between the category search request and each candidate search result is calculated by adopting a relevance determination mode aiming at the search request aiming at the search category of the proper nouns and the characteristics of the proper nouns, and the accuracy of the relevance is favorably improved.

In another embodiment of the present application, when the search intention category is a non-proper noun search, the determining the relevance between the search request and each candidate search result according to the relevance determining manner corresponding to the search intention category provided in step S130 may include:

b11, determining the text matching degree between each candidate search result and the search request;

and B12, determining the relevance of the search request and each candidate search result based on the text matching degree between each candidate search result and the search request.

Firstly, obtaining text information corresponding to the search request and each candidate search result, and if the search request provided by the user is a search request of voice information, firstly, converting the voice information into the text information. The text information corresponding to the search request may be an entire text corresponding to the search request, or may be a subfile obtained by segmenting the entire text, and accordingly, the text information corresponding to each candidate search result may also be the entire text and the subfile, and the text matching degree between the text information corresponding to the search request and the text information corresponding to each candidate search result is calculated, and the text matching degree may be calculated by various methods, such as edit distance, BM25 (binary independent model) correlation, ngram (multiple language model) correlation, and the like. And determining the correlation degree of each candidate search result and the search request according to the text matching degree corresponding to each candidate search result, wherein the text matching degree corresponding to the candidate search result can be used as the correlation degree of the candidate search result and the search request.

According to the scheme provided by the embodiment, the text information corresponding to the search request and the text information corresponding to each candidate search result are used for calculating the text matching degree, and the correlation degree between the candidate search result and the search request can be visually determined according to the text matching degree.

On the basis of obtaining the text matching degrees corresponding to the search request and each candidate search result, for the non-specific nouns, the search method provided in an optional embodiment of the present application may further include:

b21, acquiring the relevant information of the search request and the relevant information corresponding to each candidate search result;

and B22, for each candidate search result, determining the relevant information matching degree of the relevant information corresponding to the candidate search result and the relevant information of the search request.

The related information of the search request can reflect the characteristics of the search request from the side, and the related information of the candidate search result also reflects the characteristics of the candidate search result from the side, such as: the search request is coffee, the candidate search result is coffee shop, the relevance of the coffee shop and the candidate search result is low on the text, but the main business of the coffee shop is coffee selling substantially, the business information is relevant information of the coffee shop, and deep connection between the search request and the candidate search result can be mined based on the relevant information of the search request and the relevant information of the candidate search result. Based on the relevant information matching degree between the relevant information of the search request and the relevant information of the candidate search results, the consideration factor of the relevant degree between the search request and the candidate search results is expanded, the relevance between the search request and the candidate search results is prevented from being measured from a single angle, and the accuracy of the relevant degree between the search request and the candidate search results is improved.

The related information of the search request may include service information, entity information, attribute information, and the like of the search request, and the service information, entity information, and attribute information of the candidate search result are illustrated as follows: the entity information of the search request is coffee, and the candidate search result comprises: the business information of the coffee hall in the candidate search result is coffee, the business of the gymnasium is fitness, the entity of the coffee-flavored ice cream is ice cream, and the 'mug', 'ice', and the like in the candidate search result are attribute information.

And respectively carrying out relevance matching calculation on the relevant information corresponding to the search request and the relevant information corresponding to the candidate search results to obtain the relevance information matching degree of the search request and each item of relevant information of each candidate search result. The matching degree of the related information corresponding to each item of related information may be used as the correlation degree between the search request and each candidate search result, or the matching degree of the related information corresponding to each item of related information may be used as a parameter for calculating the correlation degree between the search request and the candidate search result.

In an alternative embodiment of the present application, the determining the relevance of the search request to each candidate search result based on the text matching degree between each candidate search result and the search request, provided by B21, includes:

and for each candidate search result, determining the relevance of the search request and the candidate search result based on the text matching degree and the relevant information matching degree corresponding to the candidate search result.

In the embodiment, the relevancy between the search result and each candidate search result is determined according to the text matching degree between the search request and each candidate search result and the relevant information matching degree, and is determined based on the text matching degree and the relevant information matching degree, so that not only is the matching degree of the search request and the candidate search result on the text considered, but also the matching degree of the search request and the candidate search result on the relevant information considered, and the accuracy of the relevancy between the search request and the candidate search result is further improved.

On the basis of obtaining the text matching degrees corresponding to the search request and each candidate search result, for the non-specific nouns, the search method provided in another alternative embodiment of the present application may further include:

b31, for each candidate search result, determining semantic similarity of the candidate search result to the search request.

The semantic similarity between the candidate search result and the search request can accurately represent the similarity between the semantic information of the candidate search result and the semantic information of the search request. The semantic similarity between the search request and the candidate search result can be calculated through a semantic recognition model, and the semantic recognition model can dig out the similarity between the search request and the real semantic information of the candidate search result, so that the semantic similarity between the candidate search result and the search request can be used as the correlation between the candidate search result and the search request, and the accurate representation of the correlation between the candidate search result and the search request is realized.

In the training process of the semantic recognition model, semantic information of a search request and semantic information corresponding to candidate search results are used as input of the model, semantic similarity between the search request and the candidate search results is output, the similarity can range from 0 to 1 and includes endpoint values, the closer the correlation is to 0, the lower the correlation is, and the closer the correlation is to 1, the higher the correlation is.

For example, the search request includes "breakfast", and the corresponding candidate search results include three results of a steamed bunk, a breakfast shop, and a fast food restaurant, although the matching degrees of the three candidate results in text matching and related information matching are different, in essence, the three candidate search results and the search request have higher correlation, and the semantic similarity provided by the semantic recognition model can make up for the deficiency of the text matching degree and the related information matching degree in calculating the correlation between the search request and each candidate search result, which is beneficial to improving the accuracy of the correlation between the search request and the candidate search results.

On the basis of obtaining the semantic similarity between each candidate search result and the search request, the determining the relevance between the search request and each candidate search result based on the text matching degree between each candidate search result and the search request provided in step B12 includes:

and for each candidate search result, determining the correlation degree of the search request and the candidate search result based on the semantic similarity and the text matching degree corresponding to the candidate search result.

The embodiment of the application provides the following scheme: for each candidate search result, determining the correlation degree of the search request and the candidate search result based on the semantic similarity and the text matching degree corresponding to the candidate search result, wherein the text matching degree may be determined only according to the text matching degree between the subsequent selected search result and the search request, or may be obtained based on the text matching degree and the correlation information matching degree corresponding to the candidate search result.

When the text matching degree between the candidate search result and the search request is determined according to the text matching degree and the related information matching degree corresponding to the candidate search result, in combination with the search flowchart corresponding to the non-specific nouns shown in the left part of fig. 2, the search scheme corresponding to the flowchart is as follows: when the search intention category of the search request is the search of the non-proper nouns, the search process includes three layers, the first layer (corresponding to the level 1 in fig. 2) obtains the text matching degree between the search request and the candidate search result (corresponding to the text correlation in fig. 2), and then performs consistency check on the search request and each candidate search result (corresponding to the level 2 in fig. 2), that is, obtains the relevant information matching degree between the search request and the candidate search result, and the consistency check may be performed in the following order: and finally, obtaining semantic similarity between the search request and each candidate search result (corresponding to the 3 rd floor in fig. 2) through the service information, the main body information and the attribute information, and obtaining the semantic similarity between the search request and the candidate search result through a semantic recognition model (the semantic recognition model can be an end-to-end deep semantic model in fig. 2). And determining the correlation degree between the search request and the candidate search result based on the text matching degree, the related information matching degree and the semantic similarity.

Further, weights corresponding to the text matching degree, the related information matching degree and the semantic similarity can be set, and the correlation degree between the search request and any candidate search result is obtained in a linear weighting mode according to the three parameters and the corresponding weights. And determining the correlation degree of the search request and the candidate search result through multiple factors by combining the text matching degree, the related information matching degree and the semantic similarity, so that the accuracy of the correlation degree is further improved.

An alternative way to determine the text matching degree between each candidate search result and the search request is further provided in an embodiment of the present application, which includes:

for each candidate search result, the following operations are performed, and the flowchart is shown in fig. 3:

s310, acquiring information points in the candidate search results, and acquiring at least two target fields based on the information points;

s320, determining the text similarity between the search request and each target field;

s330, determining the text matching degree between the candidate search result and the search request based on the text similarity corresponding to each target field.

Determining the text matching degree between the candidate search result and the search request, not only directly calculating the overall text matching degree between the candidate search result and the search request, but also obtaining the information point of each candidate search result, wherein the information point of the candidate search result is data representing the key point information of the candidate search result, and obtaining at least two target fields based on the information point of each candidate search result, such as: if the candidate search result is a store, the information point of the store may include a store name, a store category, an address name, a product name in the store, and the like of the store, and at least two object fields may be determined based on the information point, which may be the store name, the product name, the address name, and the like. And then aiming at each target field, obtaining the text matching degree between each target field and the search request, determining the text matching degree between the corresponding candidate search result and the search request based on the text matching degree corresponding to each target field, dividing the candidate search result into target fields with smaller length and specific information, determining the text matching degree between the candidate search result containing the target fields and the search request based on the text similarity of the target fields and the search request, and being beneficial to improving the accuracy of the text matching degree between the search request and the candidate search result.

In another alternative embodiment, the weights of the target fields corresponding to each candidate search result are different according to the scheme provided by the above embodiment. In this case, determining the text matching degree between the candidate search result and the search request based on the text similarity corresponding to each target field, which is provided in S330, may be implemented as follows:

s331, acquiring the weight of each target field;

s332, according to the weight of each target field, carrying out weighted summation on the text similarity of each target field to obtain the text matching degree between the candidate search result and the search request.

According to the scheme provided by the embodiment, when the text matching degree between the candidate search result and the search request is determined according to the text similarity of each target field, the weights of different target fields are considered, and the text matching degree between the candidate search result containing the target field and the search request can be determined in a linear weighting mode based on the weight corresponding to the target field and the text similarity, so that the accuracy of the text matching degree between the candidate search result and the search request can be further improved.

In an alternative embodiment, the step B21 of providing relevant information including at least one of service information, entity information, and attribute information, and if the relevant information includes at least two items, determining, for each candidate search result, a matching degree between relevant information corresponding to the candidate search result and relevant information of the search request may include:

c1, acquiring the weight corresponding to each item of information in the related information;

c2, determining the matching degree of each item of information in the candidate search result and the relevant information of the search request;

and C3, performing weighted summation on the matching degrees of the relevant information items based on the weights corresponding to the relevant information items, and obtaining the matching degree of the relevant information corresponding to the candidate search result and the relevant information of the search request.

The related information provided by the embodiment of the present application includes at least one of service information, entity information, and attribute information, that is, any one of the three items, or a combination of any two or three items.

When the related information comprises at least two items, determining the weight corresponding to each item of related information, and determining the matching degree of the candidate search result and the related information of the search request based on the weight. Specifically, the process of determining the matching degree of the candidate search result and the relevant information of the search request is as follows: firstly, the weight corresponding to each item of information in the related information is obtained, the weight can be obtained through learning according to historical data or cloud big data, then the related information and the corresponding weight are subjected to weighted summation, and the matching degree of the related information of each item of related information of the candidate search result and the related information corresponding to the search request is obtained.

If only one kind of relevant information is included in a candidate search result, the weight of the missing relevant information may be set to 0, and the matching degree of the relevant information corresponding to the candidate search result may be calculated. If the matching degree of the relevant information between the entity information of a certain candidate search result and the entity information of the search result is 0, no matter the size of the weight corresponding to the preset entity information, the matching degree of the relevant information is multiplied by the corresponding weight to obtain that the matching degree corresponding to the entity information is 0, and the schemes for determining the matching degrees of the relevant information corresponding to the service information and the attribute information are similar and are not repeated.

The scheme provided by the embodiment further refines the related information, and calculates the related information matching degree aiming at the refined related information, so that the accurate related information matching degree can be obtained, and the accuracy of the correlation degree determined based on the related information matching degree can be improved.

When determining the matching degree between the related information of the search result and the related information corresponding to each candidate search result, in addition to the above scheme, the following scheme may be used to obtain the related information, where the related information includes first information and second information, the first information includes at least one of service information and entity information, the second information includes attribute information, and the matching degree between the related information corresponding to the candidate search result and the related information of the search request is determined, which may be implemented in the following manner, and a flowchart of the method shown in fig. 4 includes:

s410, determining the matching degree of the relevant information corresponding to each item of relevant information;

s420, determining the priority of each candidate search result according to the matching degree of the related information corresponding to the first information, and obtaining at least two priorities and each candidate search result corresponding to each priority;

s430, aiming at each priority, determining the ordering information corresponding to each candidate search result belonging to the priority according to the relevant information matching degree corresponding to the second information of each candidate search result belonging to the priority;

s440, for each candidate recognition result, determining the matching degree of the relevant information corresponding to the candidate search result and the relevant information of the search request based on the priority and the sorting information corresponding to the candidate recognition result.

Optionally, each priority level may be preset with corresponding basic metric information, such as: the basic score, the ranking information corresponding to each candidate search result in each priority may be a ranking score, and for each candidate search result, the matching degree of the relevant information corresponding to the candidate search result may be determined according to the priority and the ranking information of the candidate search result, such as: the candidate search result is obtained by summing the basic score and the ranking score of the priority corresponding to the candidate search result, or the candidate search result is obtained by performing weight distribution on the basic score and the ranking score, and performing weighted summation on the basic score, the weighting score and the corresponding weight.

The scheme provided by the embodiment of the application provides a method for determining the matching degree of the relevant information corresponding to the candidate search results and the relevant information of the search request, the candidate search results are roughly divided through the service information or/and the entity information, then the attribute information of the candidate search results is further ranked, the detailed ranking of each candidate search result in each level is favorably obtained, and the data calculation amount in the ranking process of each candidate search result is favorably reduced.

According to the scheme provided by the embodiment, the relevance corresponding to each candidate search result is obtained, the target search result is screened from the candidate search results according to the relevance, the candidate search results with the relevance larger than the preset threshold can be used as the target search results by setting the preset threshold, the candidate search results can be ranked according to the relevance according to the preset number of the preset target search results, the candidate search results with the preset number of the candidate search results ranked in the front are used as the target search results, the expectation of a user is met, and the user experience is improved.

When ranking the candidate search results, in addition to ranking the candidate search results by using the relevance, an optional embodiment of the present application further provides the following scheme for ranking the candidate search results, specifically as follows: after obtaining the correlation degree between each candidate search result and the search request, classifying each candidate search result according to the correlation degree so as to determine the correlation degree between the candidate search result and the search request directly according to the level of the candidate search result, such as: and dividing the correlation into 2 levels which are respectively a first level and a second level, wherein the correlation corresponding to the first level is higher than the correlation corresponding to the second level, and the correlation of each divided candidate search result cannot cross the levels.

The process of ranking candidate search results based on relevance according to the embodiment of the present application is described with reference to the schematic diagram provided in fig. 5: after obtaining the correlation degree corresponding to each candidate search result of the search request, the correlation degree is characterized in a score form, and according to the score size of the correlation degree, each candidate search result is divided into a plurality of ranks, for example, each candidate search result can be divided into two ranks, namely, simply divided into two ranks, i.e., GOOD (corresponding to GOOD in fig. 5) and bad (corresponding to BED in fig. 5), and the correlation degree corresponding to each candidate search result does not cross the ranks, that is, the candidate search result with bad rank is not ranked before the candidate search result with GOOD rank (corresponding to GOOD store in fig. 5). For each gear, for each candidate search result in the gear, a smooth score of the candidate search result in the gear may also be obtained, the smooth score may be used as a ranking of the candidate search results in the gear to which the candidate search result belongs, and the candidate search result and the corresponding smooth score may also be used as input of a model for obtaining the smooth score, where the smooth score may be obtained by fusing multiple factors such as a conversion rate, a click rate, and a quality of the candidate recognition result.

According to the scheme provided by the embodiment, the candidate search results are classified into levels by utilizing the relevance between the candidate search results and the search request, the candidate search results in the levels are subjected to multi-factor joint sequencing by utilizing historical feedback information such as conversion rate, click rate and the like, factors influencing the candidate search results are comprehensively considered, the sequencing accuracy of the candidate search results is favorably improved, and the sequencing accuracy of the target search results can be improved based on the accurate sequencing of the candidate search results.

Optionally, after determining the target search result from the candidate search results in the manner provided in the above embodiment, when the target search result includes at least two target search results, determining the target search result of the user from each candidate search result and providing the target search result to the user may include:

d1, determining the ranking information of each target search result based on the relevance;

d2, providing the target search results to the user according to the ranking information.

When the target search results comprise at least two target search results, the arrangement sequence of the target search results is determined before the target search results are provided for the user.

In order to better understand and explain the scheme and the beneficial effects of the embodiment of the present application, the following describes a search method provided by the present application by using a specific example, the scheme of the present application can be applied to an application having a search function, the following example is described by taking a takeaway application as an example, a user searches information such as a product or a shop through the application, and the following example is described by taking a search request as an example of a search request for coffee:

aiming at a search request of a non-proper noun ' coffee ', corresponding candidate search results comprise an ' A coffee shop ', ' B Western fast food shop ', ' coffee ice cream ' and a body-building experience course ', firstly, the text matching degree of each candidate search result and the search request is determined, the text matching degree of the ' coffee ice cream ' in the candidate search results is the highest, secondly, the matching degree of each candidate search result and relevant information of the search request is determined, the corresponding relevant information matching degree is determined according to entity information, business information and attribute information of the candidate search results, the consistency of the business information of the ' A coffee shop ' and the ' coffee ' in the candidate search results is the highest, and finally, the semantic similarity between each candidate search result and the search request is obtained through a model, and the semantic similarity of the ' A coffee shop ' in the candidate search results is the highest. And obtaining the text matching degree, the related information matching degree and the weight corresponding to the semantic similarity, carrying out weighted summation on the text matching degree, the related information matching degree, the semantic similarity and the corresponding weight to obtain the final correlation degree of the candidate search result and the search request, wherein the correlation degree of the coffee shop is the highest in the candidate search results, and if only one candidate search result is selected as the target search result, the coffee shop is taken as the target search result of the coffee to be provided for the user.

According to the scheme provided by the embodiment of the application, when the correlation degree of the search request and the candidate search result is determined, the matching of the candidate search result and the search request on the text is determined through the text matching degree, the matching of the candidate search result and the search request on the related information is determined through the related information matching degree, the matching of the candidate search result and the search request on the actual semantics is determined through the semantic similarity, the accurate representation of the correlation between the candidate search result and the search request is realized through the matching of various information, and the accurate recommendation is conveniently carried out on the user based on the accurate representation.

Based on the same principle as the method provided by the embodiment of the present application, the embodiment of the present application further provides a search apparatus 600, as shown in fig. 6, the apparatus may include: a candidate search result obtaining module 610, a search intention category determining module 620, a relevance determining module 630, and a target search result determining module 640, wherein:

a candidate search result obtaining module 610, configured to obtain a search request of a user and candidate search results corresponding to the search request;

a search intention category determination module 620 for determining a search intention category of the search request;

a relevance determining module 630, configured to determine relevance between the search request and each candidate search result according to a relevance determining manner corresponding to the search intention category;

and the target search result determining module 640 is configured to determine a target search result of the user from the candidate search results and provide the target search result to the user based on the correlation corresponding to each candidate search result.

According to the searching device, the searching request is divided into the searching intention categories, the correlation degree determining mode corresponding to the searching intention categories of the searching request is determined, the correlation degree calculation is performed in a targeted mode according to different searching intention categories, the correlation degree between the searching request and each candidate searching result is obtained accurately according to the searching intention of a user, the target searching result provided for the user is enabled to be more consistent with the expectation of the user based on the accurate correlation degree corresponding to each candidate searching result, the matching degree of the target searching result and the expectation of the user is improved, and the use perception of the user is improved.

In an embodiment of the present application, if the search intention category is proper noun search, the relevancy determination module 630 is specifically configured to:

mining candidate proper nouns in the search request;

determining standard proper nouns corresponding to the candidate proper nouns based on the matching degree of the candidate proper nouns and the standard proper nouns in the proper noun database;

and determining the correlation degree between the search request and each candidate search result based on the matching degree of the standard proper nouns corresponding to the candidate proper nouns and each candidate search result.

Optionally, if the search intention category is non-proper noun search, the relevancy determination module 630 is specifically configured to:

determining the text matching degree between each candidate search result and the search request;

and determining the relevance of the search request and each candidate search result based on the text matching degree between each candidate search result and the search request.

Optionally, the search apparatus 600 further includes a related information matching degree determining module, specifically configured to:

acquiring relevant information of the search request and relevant information corresponding to each candidate search result;

for each candidate search result, determining the matching degree of the relevant information corresponding to the candidate search result and the relevant information of the search request;

in this embodiment, the correlation determining module 630 is specifically configured to:

Optionally, the search apparatus 600 further includes a semantic similarity determining module, configured to:

for each candidate search result, determining semantic similarity between the candidate search result and the search request;

Optionally, the correlation determination module 630 is further configured to:

for each candidate search result, the following operations are performed:

acquiring information points in the candidate search results, and acquiring at least two target fields based on the information points;

determining the text similarity between the search request and each target field;

and determining the text matching degree between the candidate search result and the search request based on the text similarity corresponding to each target field.

Optionally, the correlation determination module 630 is specifically configured to:

acquiring the weight of each target field;

and carrying out weighted summation on the text similarity of each target field based on the weight of each target field to obtain the text matching degree between the candidate search result and the search request.

Optionally, the related information includes at least one of service information, entity information, and attribute information, and if the related information includes at least two items, the related information matching degree determining module is specifically configured to:

acquiring the weight corresponding to each item of information in the related information;

determining the matching degree of each item of information in the candidate search results and the related information of the search request;

and carrying out weighted summation on the matching degrees of the relevant information based on the weights corresponding to the relevant information to obtain the matching degree of the relevant information corresponding to the candidate search result and the relevant information of the search request.

Optionally, the related information includes first information and second information, the first information includes at least one of service information and entity information, the second information includes attribute information, and the related information matching degree determining module is specifically configured to:

determining the matching degree of the relevant information corresponding to each item of relevant information;

determining the priority of each candidate search result according to the matching degree of the related information corresponding to the first information, and obtaining at least two priorities and each candidate search result corresponding to each priority;

aiming at each priority, determining the ordering information corresponding to each candidate search result belonging to the priority according to the relevant information matching degree corresponding to the second information of each candidate search result belonging to the priority;

and for each candidate identification result, determining the matching degree of the relevant information corresponding to the candidate search result and the relevant information of the search request based on the priority corresponding to the candidate identification result and the sorting information.

Optionally, the target search result includes at least two, and the target search result determining module 640 is specifically configured to:

determining ranking information of each target search result based on the relevance;

and providing each target search result to the user according to the sequencing information.

The search apparatus of the embodiment of the present application can execute the search method provided by the embodiment of the present application, and the implementation principles thereof are similar, the actions executed by each module and unit in the search apparatus in each embodiment of the present application correspond to the steps in the search method in each embodiment of the present application, and for the detailed functional description of each module of the search apparatus, reference may be specifically made to the description in the corresponding search method shown in the foregoing, and details are not repeated here.

Based on the same principle as the method shown in the embodiments of the present application, there is also provided in the embodiments of the present application an electronic device, which may include but is not limited to: a processor and a memory; a memory for storing a computer program; a processor for executing the search method according to any of the alternative embodiments of the present application by calling a computer program. Compared with the prior art, the search method provided by the application has the advantages that the search request is divided into the search intention categories, the relevance determining mode corresponding to the search intention categories of the search request is determined, the relevance calculation is performed on different search intention categories, the relevance between the search request and each candidate search result can be accurately obtained, the target search result provided for a user can better accord with the expectation of the user on the basis of the accurate relevance corresponding to each candidate search result, and the matching degree of the target search result and the expectation of the user is improved.

In an alternative embodiment, an electronic device is provided, as shown in fig. 7, the electronic device 4000 shown in fig. 7 may be a server, including: a processor 4001 and a memory 4003. Processor 4001 is coupled to memory 4003, such as via bus 4002. Optionally, the electronic device 4000 may further comprise a transceiver 4004. In addition, the transceiver 4004 is not limited to one in practical applications, and the structure of the electronic device 4000 is not limited to the embodiment of the present application.

The Processor 4001 may be a CPU (Central Processing Unit), a general-purpose Processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other Programmable logic device, a transistor logic device, a hardware component, or any combination thereof. Which may implement or perform the various illustrative logical blocks, modules, and circuits described in connection with the disclosure. The processor 4001 may also be a combination that performs a computational function, including, for example, a combination of one or more microprocessors, a combination of a DSP and a microprocessor, or the like.

Bus 4002 may include a path that carries information between the aforementioned components. The bus 4002 may be a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus 4002 may be divided into an address bus, a data bus, a control bus, and the like. For ease of illustration, only one thick line is shown in FIG. 7, but this is not intended to represent only one bus or type of bus.

The Memory 4003 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical Disc storage, optical Disc storage (including Compact Disc, laser Disc, optical Disc, digital versatile Disc, blu-ray Disc, etc.), a magnetic Disc storage medium or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited to these.

The memory 4003 is used for storing application codes for executing the scheme of the present application, and the execution is controlled by the processor 4001. Processor 4001 is configured to execute application code stored in memory 4003 to implement what is shown in the foregoing method embodiments.

Among them, electronic devices include but are not limited to: mobile terminals such as mobile phones, notebook computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and the like, and fixed terminals such as digital TVs, desktop computers, and the like. The electronic device shown in fig. 7 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present application.

The server provided by the application can be an independent physical server, can also be a server cluster or distributed system formed by a plurality of physical servers, and can also be a cloud server for providing basic cloud computing services such as cloud service, a cloud database, cloud computing, cloud functions, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN (content delivery network) and big data and artificial intelligence platforms. The terminal may be, but is not limited to, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and the like. The terminal and the server may be directly or indirectly connected through wired or wireless communication, and the application is not limited herein.

The present application provides a computer-readable storage medium, on which a computer program is stored, which, when running on a computer, enables the computer to execute the corresponding content in the foregoing method embodiments.

It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and may be performed in other orders unless explicitly stated herein. Moreover, at least a portion of the steps in the flow chart of the figure may include multiple sub-steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed alternately or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

It should be noted that the computer readable medium mentioned above in the present application may be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present application, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In this application, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.

The computer readable medium may be embodied in the electronic device; or may exist separately without being assembled into the electronic device.

The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to perform the methods shown in the above embodiments.

According to an aspect of the application, a computer program product or computer program is provided, comprising computer instructions, the computer instructions being stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions to cause the computer device to perform the search method provided in the various alternative implementations described above.

Computer program code for carrying out operations for aspects of the present application may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The modules described in the embodiments of the present application may be implemented by software or hardware. Where the name of a module does not in some cases constitute a limitation on the module itself, for example, the search intention category determination module may also be described as a "search intention category module that determines a search request.

The above description is only a preferred embodiment of the application and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the disclosure herein is not limited to the particular combination of features described above, but also encompasses other arrangements formed by any combination of the above features or their equivalents without departing from the spirit of the disclosure. For example, the above features may be replaced with (but not limited to) features having similar functions disclosed in the present application.

Claims

1. A method of searching, comprising:

determining a search intent category of the search request;

determining the correlation degree of the search request and each candidate search result according to a correlation degree determination mode corresponding to the search intention category;

and determining a target search result of the user from each candidate search result and providing the target search result to the user based on the corresponding correlation of each candidate search result.

2. The method of claim 1, wherein if the search intention category is proper noun search, the determining the relevance of the search request to each of the candidate search results according to the relevance determination manner corresponding to the search intention category comprises:

mining candidate proper nouns in the search request;

determining a standard proper noun corresponding to the candidate proper noun based on the matching degree of the candidate proper noun and the standard proper noun in a proper noun database;

and determining the correlation degree between the search request and each candidate search result based on the matching degree of the standard proper noun corresponding to the candidate proper noun and each candidate search result.

3. The method of claim 1, wherein if the search intention category is non-proper noun search, the determining the relevance of the search request to each of the candidate search results according to the relevance determination manner corresponding to the search intention category comprises:

4. The method of claim 3, further comprising:

determining the relevance of the search request to each candidate search result based on the text matching degree between each candidate search result and the search request comprises:

5. The method of claim 3 or 4, further comprising:

for each candidate search result, determining semantic similarity of the candidate search result and the search request;

6. The method of claim 3, wherein determining a text match between each of the candidate search results and the search request further comprises:

for each candidate search result, the following operations are performed:

7. The method of claim 6, wherein determining the text matching degree between the candidate search result and the search request based on the text similarity corresponding to each target field comprises:

acquiring the weight of each target field;

8. The method according to claim 4, wherein the related information includes at least one of service information, entity information, and attribute information, and if the related information includes at least two items, the determining, for each candidate search result, a matching degree of the related information corresponding to the candidate search result with the related information of the search request includes:

determining the matching degree of each item of information in the candidate search result and the related information of the search request;

9. The method of claim 4, wherein the related information includes first information and second information, the first information includes at least one of service information and entity information, the second information includes attribute information, and the determining the degree of matching between the related information corresponding to the candidate search result and the related information of the search request includes:

and for each candidate identification result, determining the matching degree of the relevant information corresponding to the candidate search result and the relevant information of the search request based on the priority and the sorting information corresponding to the candidate identification result.

10. The method of claim 1, wherein the target search results comprise at least two, and wherein determining the target search result of the user from the candidate search results and providing the target search result to the user comprises:

determining ranking information of the target search results based on the relevance;

and providing the target search results to the user according to the sequencing information.

11. A search apparatus, comprising:

a relevancy determining module, configured to determine relevancy between the search request and each candidate search result according to a relevancy determining manner corresponding to the search intention category;

and the target search result determining module is used for determining a target search result of the user from each candidate search result and providing the target search result to the user based on the corresponding correlation of each candidate search result.

12. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of any one of claims 1-10 when executing the program.

13. A computer-readable storage medium, characterized in that a computer program is stored on the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method of any one of claims 1-10.